Oxygen extraction fraction (OEF) is an important cerebral oxygen metabolic parameter that provides critical information for understanding the pathophysiology of cerebrovascular disease, predicting tissue infarction and patient outcome, and may identify patients who will benefit from therapeutic intervention. It has been demonstrated that elevation of oxygen extraction fraction (OEF) in brain tissue portends a 6-7 fold increased risk of subsequent stroke in patients with chronic atherosclerotic carotid artery occlusion. Moreover, during acute ischemic stroke, misery perfusion, marked by an elevation of OEF and decreased blood flow, has been used as an indicator of at-risk tissue that is a target of therapeutic treatments. On the other hand, indiscriminate treatment without screening for at-risk tissue may lead to serious treatment- related complications. Therefore, a reliable and non-invasive measure of OEF that can be obtained with a routine clinical setup will have a great impact on appropriate patient diagnosis and treatment. Despite its importance, cerebral oxygenation can only be quantified using 15O positron emission tomography (PET). However, the requirement of an onsite cyclotron has limited 15O PET measurements to only a few medical centers around the world. In the past decade, significant technical progress has been made towards non-invasive quantitative OEF measurements using magnetic resonance imaging (MRI). Though absolute OEF measurements have been successfully obtained in healthy human subjects, the clinical utility of these methods remains minor due to a host of unaddressed issues that are particularly challenging in patient studies. Moreover, neither global nor regional OEF validation has yet been performed in patients. In this study, an integrated PET/MR system which allows truly simultaneous PET and MR image acquisition will be utilized for a direct comparison between MR and PET OEF without potential confounding factors associated with sequential imaging. The overall objectives of this project are (1) to develop a novel MR OEF measurement that is rapid and robust to meet the need of clinical studies; and (2) to validate the MR OEF measurements using simultaneous O15 PET measurements in patients with focal pathology. Since MR is widely available, the success of this project will permit the acquisition of PET comparable OEF measurements in many patients, and pave the way for future integration of this method into time-sensitive patient management decisions. Furthermore, the technical developments from this proposal-- minimizing MR imaging artifacts, improving measurement efficiency, and synergizing PET and MR scans--can be generalized to many other neuroimaging studies.